Lecture 4

Capital Asset Pricing Model

(Chap. 11)

Riccardo Zago

ESCP Business School

2024 / 2025
Introduction

Determine the appropriate risk premium (Capital Asset Pricing

Model / CAPM)
I Tobin (1958), Jack Treynor (1962), William Sharpe (1964),
John Lintner (1965), Jan Mossin (1966).

James Tobin William Sharpe Harry Markowitz

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 2 / 27

Outline

1 Efficient portfolio and required returns (Chap. 11.6)

2 Capital Asset Pricing Model (Chap. 11.7)
3 Determining the risk premium (according to the CAPM) (Chap.
11.8)

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 3 / 27

A. Efficient portfolio and required returns

To start, we ask a basic question:

How can we improve the Sharpe ratio of our portfolio?

Example: investors wish to add the security A to their portfolio

P (any portfolio).
I Can we determine whether this security improves the Sharpe
ratio of their portfolio? If yes, how?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 4 / 27

Improving the Sharpe ratio of a portfolio: example
Example of the set of possible portfolios of P and the security A

Purchase of A is financed by borrowing at the risk-free rate.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 5 / 27
Improving the Sharpe ratio of a portfolio: example
(con’t)
Investors who hold the portfolio P borrow a “small” quantity δ
at the risk-free rate and invest this amount in A. The return of
their portfolio changes by: ∆R = δ (E (RA ) − rf ) .
The risk of their portfolio changes by (using a first-order Taylor
expansion, see appendix):
q
∆σ = σp2 + δ 2 σA2 + 2δσP σA ρA,P − σP
σP σA ρA,P
≈δ = δσA ρA,P
σP
This result is intuitive: the contribution to the risk of their
portfolio depends not only of the risk of the new security A but
also on the correlation with their existing portfolio!
I Not convinced? Start with the general expression of the
variance of a portfolio (Eq. 11.13 in the textbook).
Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 6 / 27
Improving the Sharpe ratio of a portfolio: example
(con’t 2)
A similar increase of the risk of the existing portfolio (∆σ) via
leverage would increase the return by:

E (Rp ) − rf
∆σ .
σP
Hence, it is optimal to add security A if and only if:

E (Rp ) − rf
δ [E (RA ) − rf ] ≥ δσA ρA,P .
σP
Or:
σA
E (RA ) ≥ rf + ρA,P [E (Rp ) − rf ] = rf + βA,P [E (Rp ) − rf ] .
σP

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 7 / 27

To sum up

Investors should add security A to their existing portfolio P only

if
E (RA ) ≥ rf + βA,P [E (Rp ) − rf ] .

I Remark: If the return is lower, the best strategy is to short the

security.
Hence, we define the required return of an asset as:

ri = rf + βi,P [E (Rp ) − rf ] .

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 8 / 27

Example: required return of a new investment

Omega: expected return of 15% and volatility of 20%.

Treasury bills (risk-free) pay 3%.
REIT (real estate investment trust): expected return of 9% and
volatility of 35%. Correlation with Omega is 10%.

Should we invest in the REIT if we already hold Omega?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 9 / 27

B. Capital Asset Pricing Model

Reminder: Markowitz (1952) + Tobin (1958) → any investor

holds a combination of a risk-free investment and the
super-efficient portfolio.
How can we extend this result to a theory of market equilibrium?
⇒ Let us add a couple of hypotheses to the previous result.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 10 / 27

CAPM: hypotheses

Starting point: Markowitz (1952) / Tobin (1958).

+
Investors hold homogeneous beliefs about expected returns,
volatilities, and correlations of all financial assets.
The market is in equilibrium.
All investors hold an efficient portfolio, that is a portfolio which
offers the highest expected return for a given level of volatility.
⇒ The super-efficient portfolio is unique.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 11 / 27

Supply, demand, and the efficiency of the market
portfolio

Reminder: homogeneous beliefs → the investors all invest in the

same super-efficient portfolio.
I All risky assets are held by all investors in the same proportions
(within the risky portfolio).

But investors put different weights on the super-efficient

portfolio and the risk-free asset depending on their risk aversion.
Does this permit to pin down the composition of the
super-efficient portfolio?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 12 / 27

Supply, demand and the efficiency of the market
portfolio: example

Camille has identified the efficient portfolio and decides to hold

it.
She holds 10 000 EUR in stocks of Capgemini and 5 000 EUR in
stocks of Sanofi.
Tom, more risk averse, only holds 2 000 EUR invested in Sanofi
stocks.
1 Assuming that his portfolio is also efficient, what is his
investment in Capgemini stocks?
2 If all investors hold the efficient portfolio, what can we say
about the total market capitalisation of Capgemini relative to
the one of Sanofi?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 13 / 27

The efficiency of the market portfolio

It is easy to generalize the preceding example to any market.

I Investors all hold the same efficient portfolio of risky assets
(same weights).
I Equilibrium between supply and demand → efficient portfolio
contains all risky assets, weighted by their market capitalisation.
In equilibrium, the market portfolio is the super-efficient
portfolio/tangency portfolio.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 14 / 27

The Capital Market Line

The equation of the efficient frontier is then given by:

E (RM ) − rf
E (Rp ) = rf + σP .
σM
⇒ Equation of the Capital Market Line (CML)
All portfolios on the CML are made up of the market portfolio
and the risk-free asset!

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 15 / 27

Capital Market Line: example

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 16 / 27

C. Determining the risk premium

Reminder: we are still trying to pin down an equilibrium relation

for the returns on individual assets.
Up to now, we know:
I The required return on an asset (defined relative to any existing
portfolio).
I The composition of the portfolios on the efficient frontier (CML)
+ the composition of the portfolios held by each investor.
It is thus straightforward to establish the relation between the
required returns of an asset and the different market parameters!

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 17 / 27

Expected returns, required returns and market
portfolio

Special feature of any efficient portfolio: the expected return of

all assets is equal to the required return (otherwise . . . )!
And yet, each investor holds the market portfolio! Thus, in
equilibrium:

E (Ri ) = ri = rf + βi × [E (Rm ) − rf ] .
| {z }
Risk premium of security i

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 18 / 27

Market risk and beta

The previous equation shows the “beta” of an asset with respect

to the market portfolio:
z }| {
σRi × Corr (Ri , Rm ) Cov (Ri , Rm )
βi = = .
σRm Var (Rm )
Volatility of asset i that
is common with the market

The required return of an asset depends on the sensitivity of

that asset to the market risk.
I It is thus independent from the idiosyncratic risk!

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 19 / 27

Example 1: expected return on a stock

The risk-free rate is 4%.

The market portfolio has an expected return of 10% and a
volatility of 16%.
Stock A has a volatility of 16% and a correlation with the
market portfolio of 33%.

1 What is the beta of stock A and what is its expected return?

2 What is the portfolio on the CML which has the same risk as
Stock A and what is its expected return?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 20 / 27

Example 2: expected return with negative beta

The beta of the stock B is -0.3.

The risk-free rate is 4% and the expected return on the market
portfolio is 10%.
What is the expected return according to the CAPM?
What do you think about that?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 21 / 27

Security Market Line

The CAPM highlights a linear relation between the beta of a

security and the expected return.
This straight line—called the Security Market Line—passes
through the risk-free asset (β = 0) and the market portfolio
(β = 1).
In equilibrium, all assets are on the SML.
I But no asset is on the CML!!

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 22 / 27

CML and SML: illustration

Capital Market Line (CML)

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 23 / 27
CML and SML: illustration (con’t)

Security Market Line (SML)

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 24 / 27
The beta of a portfolio

P
The beta of a portfolio with return Rp = i xi Ri is given by:
P
Cov (Rp , Rm ) Cov ( i xi Ri , Rm )
βp = =
Var (Rm ) Var (Rm )
X Cov (Ri , Rm ) X
= xi = xi βi .
i Var (Rm ) i

The beta of a portfolio is equal to the weighted average of the

betas of its constituents.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 25 / 27

Example: expected return of a portfolio

The betas of Danone (BN.PA) and Ubisoft (UBI.PA) are

respectively 0.5 and 1.8.
The risk-free rate is 4%.
The expected return on the market portfolio is 10%.
According to the CAPM, what is the expected return of an
equally-weighted portfolio of these two stocks?

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 26 / 27

Summary: what to remember from the CAPM?

The market portfolio is the only efficient portfolio. Hence, the

highest expected return for a given level of volatility is achieved
by combining the market portfolio with the risk-free asset
(borrowing or lending).
The risk premium of any asset is proportional to its beta with
respect to the market. There is a linear relation between the
expected return of an asset and its (systematic) risk called the
SML.
Remark: the hypotheses of the CAPM are quite restrictive.
Nevertheless the CAPM is used widely to estimate the expected
returns of assets or the cost of capital of projects.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 27 / 27

Appendix: portfolio volatility approximation
In Section 1 of the slides, I claim that the approximation below holds:
q
∆σ = σp2 + δ 2 σA2 + 2δσP σA ρA,P − σP
σP σA ρA,P
≈δ .
σP

To prove this result, define the function f as

q
f (δ) := σp2 + δ 2 σA2 + 2δσP σA ρA,P .

From Taylor’s theorem, it follows that for δ close to zero, we have

f (δ) ≈ f (0) + f 0 (0)δ.
Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 1/2
Appendix: portfolio volatility approximation (con’t)

Moreover, it is easy to check that

δσA2 + σP σA ρA,P
f 0 (δ) = q ,
σp2 + δ 2 σA2 + 2δσP σA ρA,P

and, at δ = 0, f and its first derivative simplify to f (0) = σP and

σ σ ρ
f 0 (0) = P σAP A,P . Thus, Taylor’s theorem implies that
q σP σA ρA,P
σp2 + δ 2 σA2 + 2δσP σA ρA,P ≈ σP + δ.
σP
The approximation on the previous slide then follows almost
immediately.

Riccardo Zago Lecture 2 & 3: Optimal Portfolio Choice 2/2